Figures & data
FIGURE 1 Assignment matrix representation. Note that its dimension is N × (M + 1). First column is added, for special case where a blob is not assigned to none track (“null track”).
FIGURE 2 Optimal assignment in a hypothetical situation.
FIGURE 3 Information necessary for each blob in order to achieve incrementality.
FIGURE 4 Hill Climbing for the blobs-to-tracks assignment problem.
FIGURE 5 Simulated annealing for blobs-to-tracks assignment problem.
FIGURE 6 Tabu Search for blobs-to-tracks assignment problem.
FIGURE 7 The OpenCV visual surveillance algorithms.
TABLE 1 Tracking Algorithms Used in the Evaluation Section
FIGURE 8 Scenes of the datasets used in the experimentation: (a) SQUASH; (b) BOAT; (c) HANDBALL; (d) CORRIDOR; and (e) PETS2002. The original sequence is depicted in the first row and their respective foreground images are in the second row.
